A study of the electronic structure of ultrathin AlAs layers buried in GaAs(100) and their interfaces is presented. Al L2,3 soft-X-ray-emission (SXE) spectra from the AlAs layers were measured. The spectra show distinct thickness-dependent features, which are reproduced using ab initio calculations.
A beam/adhesive-layer model is developed. For this model a closed-form solution method applicable to arbitrary boundary conditions is presented. This enables the solution of a large number of practical problems which may be statically indeterminate. The stress state in the adhesive layer and the adherends of the beam/adhesive-layer model is also scrutinized. The method is exemplified in an analysis of the end-loaded split (ELS) specimen, commonly used to determine fracture energies of adhesive layers. The effect of the flexibility of the adhesive layer on the energy release rate and the critical crack length for stable crack growth is examined. Both symmetric and unsymmetric ELS-specimens are studied.
The valence photoemission spectra of alkali metals exhibit multiple plasmon satellite structure. The calculated spectral functions within the GW approximation show only one plasmon satellite at too large binding energy. In this Letter we use the cumulant expansion approach to obtain the spectral functions of Na and Al from ab initio calculations including the effects of band structure. The GW spectral functions are dramatically improved and the positions of the multiple plasmon satellites are in very good agreement with experiment while their intensities cannot be explained from intrinsic effects only.
We present ab initio calculations of the energy loss spectra and plasmon dispersions of the alkali metals Na, K, Rb, and Cs within the random phase approximation, including band structure effects and core electrons as well as local field effects, due to inhomogeneity in the induced charge density. Band structure effects are found to cause a negative dispersion in Cs.
The spectral functions of highly correlated systems Gd and NiO have been calculated within the GW approximation. In the case of Gd, satellites are found at the positions of the experimental main peaks. However, some quasiparticle weight remains at the local density eigenvalue but its weight is unusually small (0.3). In NiO, satellite structure is observed at around -10 eV when starting from the local density Hamiltonian. It is found, however, that the satellite intensity is reduced towards self-consistency and almost disappears. © 1996 The American Physical Society.
The magnetic moment in a solid is usually associated with the electron spins but there is an additional contribution due to the orbital motion of the electrons. For a finite system such as an atom or molecule the orbital moment can be readily calculated. However, for a periodic system the formula used for finite systems becomes ill-defined due to the presence of the position operator. In the last decade a modern theory of orbital magnetization that allows for a rigorous calculation of the magnetic moment of periodic crystals has been developed. This article provides a survey of the theoretical development of this new topic as well as recent, albeit a few, applications of the new formula to real materials. Although the original theory was worked out for non-interacting systems, there has been recent progress in the theory of orbital magnetic moment of interacting electrons in solids. To include the effects of electron-electron interactions two approaches have been proposed, one based on current spin density functional theory and another on the many-body Green's function method. The two approaches are very different but both methods provide convenient yet rigorous means of including the effects of exchange and correlations beyond the commonly used local density approximation of density functional theory.
Photoemission and inverse angle-resolved photoemission spectra are presented for Cd(0001). The data are interpreted in terms of interband transitions, density-of-states effects, and excitations of surface states.
After two decades since the discovery of ferromagnetism in manganese-doped gallium arsenide, its origin is still debated, and many doubts are related to the electronic structure. Here we report an experimental and theoretical study of the valence electron spectrum of manganese-doped gallium arsenide. The experimental data are obtained through the differences between off- and on-resonance photo emission data. The theoretical spectrum is calculated by means of a combination of density-functional theory in the local density approximation and dynamical mean field theory, using exact diagonalization as impurity solver. Theory is found to accurately reproduce measured data and illustrates the importance of correlation effects. Our results demonstrate that the manganese states extend over a broad range of energy, including the top of the valence band, and that no impurity band splits-off from the valence band edge, whereas the induced holes seem located primarily around the manganese impurity.
Photoemission line shapes can be used to extract important information about ground-state and low-energy properties. This is illustrated for the electron–phonon interaction in C60 molecules, the occupancy and hybridization of the 4f-level in Ce compounds, the effect of the Cu–O network on core-level photoemission in cuprates and the properties of one-dimensional conductors.
Mixed mode testing of adhesive layer is performed with the Mixed mode double Cantilever Beam specimen. During the experiments, the specimens are loaded by transversal and/or shear forces; seven different mode mixities are tested. The J-integral is used to evaluate the energy dissipation in the failure process zone. The constitutive behaviour of the adhesive layer is obtained by a so called inverse method and fitting an existing mixed mode cohesive model, which uses a coupled formulation to describe a mode dependent constitutive behaviour. The cohesive parameters are determined by optimizing the parameters of the cohesive model to the experimental data. A comparison is made with the results of two fitting procedures. It is concluded that the constitutive properties are coupled, i.e. the peel and shear stress depend on both the peel and shear deformations. Moreover, the experiments show that the critical deformation in the peel direction is virtually independent of the mode mixity.
The Double Cantilever Beam (DCB) specimen is a common test geometry for testing of mode I fracture properties of adhesive joints. However, when unbalances are introduced to the adherends, the adhesive layer is loaded in a combination of peel (mode I) and shear (mode II). In this work the unbalanced DCB-specimen is studied by the use of the beam/adhesive layer (B/A) model, in which the adherends are considered as beams and the adhesive layer as a generalised spring media. The effect of the thickness of adhesive layer together with the effect of the geometrical and material unbalances is analysed. The result of the B/A model is compared to the continuum model through FE-simulations. Finally, an unsymmetric DCB-specimen is dimensioned for mixed mode testing of adhesive layer.
The detailed nature of electronic states mediating ferromagnetic coupling in dilute magnetic semiconductors, specifically (Ga,Mn)As, has been an issue of long debate. Two confronting models have been discussed emphasizing host band vs. impurity band carriers. Using angle resolved photoemission we show that the electronic structure of the (Ga,Mn)As system is significantly modified from that of GaAs throughout the valence band. Close to the Fermi energy, the presence of Mn induces a strong mixing of the bulk bands of GaAs, which results in the appearance of a highly dispersive band in the gap region of GaAs.
For Mn concentrations above 1% the band reaches the Fermi level, and can thus host the delocalized holes needed for ferromagnetic coupling. Overall, our data provide a firm evidence of delocalized carriers belonging to the modified host valence band.
This thesis is devoted to studies of various electronic properties which can be extracted from photoemission spectroscopy. Both weakly and strongly correlated systems are discussed, ranging from free electron metal surfaces to strongly correlated copper oxide compounds. In order to study such a variety of systems, different calculational schemes have to be employed. This work is mainly based on band structure calculations using the local density approximation. However, for strongly correlated systems the need for model Hamiltonians is crucial. Subjects discussed are: * inverse photoemission spectra from the Cd(0001) surface obtained by a multiple scattering formalism. The results are in good agreement with recent experimental data. Especially, the one-step model is used to demonstrate image states excitations by secondary beams. * artificial structures, such as semiconductor heterojunctions and short-period superlattices, which are of vital importance for modern electronic devices. A method for tuning the band offset at semiconductor interfaces is proposed; by varying (n,m) for the GaAs/(AlAs)n(GaAs)m interface, the magnitude of the valence band offset can be designed. The metallic superlattice Mo/V was also studied, with emphasis on the application of hydrogen storage. * Schottky barrier heights. Final state screening effects can influence the Schottky barrier height, as deduced from measured core level shifts. According to our calculations, such effects may be of the order 0.1-0.2 eV. * chemical shifts in Cu2O, CuO and NaCuO2, where the Cu atom is formally mono-, di- and trivalent. We find similar chemical shifts between Cu atoms in YBa2Cu3O6.5 as between the Cu atoms in the model compounds. This suggests that different Cu atoms in YBa2Cu3O6.5 are formally mono-, di- and trivalent. The Cu 2p core level spectra were calculated using the Anderson impurity model, with parameters obtained from ab initio calculations. The results are related to the formal valence of Cu, and can be explained by a charge counting argument.
Using a recently developed Green's function formalism, we have calculated the spin-wave spectra and dispersions in Ni and Fe. For Ni(100), the dispersion exhibits two branches as observed experimentally. The calculated higher optical branch is found to be too high in energy when the standard local density approximation band-structure is used but a very good agreement with the measured dispersion is obtained when the exchange splitting is reduced, to correspond to the experimental value of the exchange splitting. We also found a double branch along Ni(111) which is not observed experimentally. For Fe, the calculated dispersion surprisingly exposes a gap midway along Γ-N in disagreement with experimental data. However, an analysis of the temperature-dependent magnetization has predicted a similar gap at the same wave vector, supporting the present calculations.
We have calculated the plasmon-width dispersion of Li within the random-phase approximation including the full band structure. The dispersion is found to be negative for small momentum transfers, which is in agreement with experiment. This anomaly is due to the band structure. The energy-loss spectra are found to exhibit a small peak at low energy which can be traced back to the gap at the zone boundaries. This confirms the interpretation of the peak as a so called zone-boundary collective state.
A Green’s function formalism for calculating spin-wave excitations is developed for practical calculations and tested for real solids. The mapping to the Heisenberg Hamiltonian commonly used in spin-wave calculations is avoided, making the formalism suitable for both localized and itinerant magnetic systems. To test the formalism, we have calculated the spin-wave spectra and dispersions of ferromagnetic Fe and Ni. The results prove to be in very good agreement with experiment and some novel features are predicted.
The exchange-correlation hole and potential of the homogeneous electron gas have been investigated within the random-phase approximation, employing the plasmon-pole approximation for the linear density response function. The angular dependence as well as the time dependence of the exchange-correlation hole are illustrated for a Wigner-Seitz radius rs=4 (atomic unit). It is found that there is a substantial cancellation between exchange and correlation potentials in space and time, analogous to the cancellation of exchange and correlation self-energies. Analysis of the sum rule explains why it is more advantageous to use a noninteracting Green function than a renormalized one when calculating the response function within the random-phase approximation and consequently the self-energy within the well-established GW approximation. The present study provides a starting point for more accurate and comprehensive calculations of the exchange-correlation hole and potential of the electron gas with the aim of constructing a model based on the local density approximation as in density functional theory.
The authors study the chemical shift of the Cu 2p core level for Cu 2O, CuO and NaCuO2, where Cu is formally monovalent, divalent and trivalent, respectively. They find that the binding energy increases with the valence, as expected, although the variation in the number of 3d electrons is small and the trivalent Cu atom is found to have a smaller net positive charge than the monovalent or divalent Cu. The origin of the calculated chemical shifts is analysed and the relation with the chemical valence is discussed.
We have studied the Cu-2p core level photoemission spectrum of a variety of cuprates, mainly focusing on the chemical shift and the shape of the leading peak. The spectra are calculated using the Anderson impurity model and we obtain a very good agreement with the experimental data. We find that the shape of the leading peak depends crucially on the structure of the Cu-O network. The main peak turns out to be quite narrow if the network consists of Cu-O-Cu bond angels of the order of 90°. On the other hand, if the Cu-O atoms are arranged with bond angles of approximately 180°, the main peak becomes substantially broader and contains a rather complicated structure. However, in some cases it is not sufficient only to consider the Cu-O network because interactions with other atoms are also important. In the model compounds Cu2O, CuO and NaCuO2, where Cu is formally monovalent, divalent and trivalent, respectively, we find that the number of 3d electrons is rather similar. Nevertheless, the binding energy increases with the valence as expected from chemical intuition. The spectra exhibit a large variation in the strength of the d9-like satellite and in the width of the main line. We, furthermore, study the chemical shift of three inequivalent Cu atoms in YBa2Cu3O6.5, and compare the results with the model compounds, which suggests that the different Cu atoms in YBa2Cu3O6.5 have formal valences of approximately one, two and three. These findings are analyzed and related to the formal valence.
We study the chemical shift of three different Cu atoms in YBa2Cu3O6.5, and compare the results with the formally mono-, di-, and trivalent model compounds Cu2O, CuO, and NaCuO2. We find similar chemical shifts between Cu in the empty chains, Cu in the planes, and Cu in the intact chains as between the Cu atoms in the model compounds. This suggests that the different Cu atoms in YBa2Cu3O6.5 have formal valences of approximately 1,2, and 3. © 1992 The American Physical Society.
We have studied the Cu core-level photoemission spectra in the Anderson impurity model for cuprates with different Cu-O networks, dimensionalities (zero, one, two, and three) and Cu valencies (two and three). We focus on the shape of the leading peak and obtain very good agreement with the experimental data. We show how the shape of the spectrum is related to the valence electronic structure and the Cu-O network but also that other atoms can play a role.
We calculate the shape of the 2p core level photoemission spectrum for Cu2O, CuO and NaCuO2, where Cu is formally mono-, di- and trivalent, respectively. Although the number of 3d electrons is similar in the three compounds, there is a large variation in the strength of the d9-like satellite and in the width of the main line. We relate the differences between Cu2O and CuO to the difference in the valence, while for NaCuO2 the detailed form of the hopping matrix elements is also found to be important. © 1992 IOP Publishing Ltd.
The electronic structure at the interface between bulk GaAs(001) and short-period superlattices of (AlAs)n(GaAs)m has been calculated using ab initio pseudopotential techniques. The results show that the valence band offsets at such interfaces are very similar to those obtained experimentally for random alloy systems, but superior transport properties are anticipated for the ordered systems.
In this Letter we stress the importance of final-state effects in photoelectron spectroscopy. In particular, we address the problem of Schottky-barrier formation, as studied via core-level shifts in photoemission. We have calculated the shift of the core-level distribution when a semiconductor surface is covered with a metal, using a wave-vector-dependent image-screening model. We conclude that final-state effects, which are generally neglected in this context, are in fact quite important. This conclusion is supported by experimental observations reported in the literature. © 1991 The American Physical Society.
We have used an image screening model to calculate the shift of the Ga(3d−1) core-level distribution when a GaAs surface is covered with metal. Using the GaAs bulk dielectric constant we find a shift of 0.1 (0.18) eV with a mean free path of 25 (8) Å. These shifts are comparable to the measured ones,1 indicating that final-state (image) screening represents an important part of core-level shifts in the vicinity of the interface.
We have studied the core level photoemission spectra of the two-dimensional cuprates Sr2CuO2Cl2, Sr2CuO2Br2, Ca2CuO2Cl2, Bi2Sr2CaCu2O8+δ, and Nd2CuO4, with particular focus on the screening response to core-hole creation in the Cu−2p3/2 level. The influence of the apex positions on the shape of the so-called main line is investigated, and found to be weak. Additionally, an Anderson impurity model was used to fit the shape of the main lines, obtaining good agreement with the data from Nd2CuO4. For the other compounds, while the energy spread of the two screening channels (local and nonlocal) is well reproduced, the theory underestimates the width of the nonlocally screened feature. The shapes of the main lines are discussed in detail.
Ultrathin AlAs(100) layers of 1-, 2-, and 5-ML thickness buried in GaAs are investigated by ab initio calculations. Unique experimental soft-x-ray emission spectra are explained in terms of interface effects and changes with layer thickness are found in the density of states. Only the central layer in the 5-ML geometry is bulklike. A valence-band offset of 0.53 eV is also found for this structure, while no offset exists in the 1- and 2-ML cases. Very good agreement is achieved between theory and experiment.
Calculations of soft-x-ray emission spectra from Si(100) layers buried in GaAs are reported. The local densities of states for Si in As and Ga sites are found to be very different. By comparison with experimental data, this difference allows us to determine the relative amounts of Si in the two types of sites. In the case of a single Si layer we find that 63 (±5)% of the buried atoms are in Ga sites.
Calculations of the soft x-ray emission spectrum have been carried out on a buried Si(100) layer in GaAs and compared with experimental data. We find that Si occupies Ga- as well as As-sites and that the local density of states is different for these two cases.
As adsorption and interaction with InP(110) is investigated by means of total-energy minimization calculations. We find that the As-P exchanged configuration has higher energy than the As/InP(110) epitaxially continued layer structure (ECLS), for all types of As reservoirs considered. In the presence of an additional As layer in ECLS (i.e., 1.5 ML adsorbed As), however, the exchanged configuration is only slightly higher in energy than the nonexchanged one. We conclude that the As-P exchange process is energetically unfavorable, and should in any case not be complete at room temperature, as suggested in a recent report. Our conclusion is supported by results of a photoemission study, including As adsorption, desorption, and redeposition, according to which the process is nonreversible.
Although cuprate high-temperature superconductors were discovered already in 1986 the origin of the pairing mechanism remains elusive. While the doped compounds are superconducting with high transition temperatures T-c, the undoped compounds are insulating due to the strong effective Coulomb interaction between the Cu 3d holes. We investigate the dependence of the maximum superconducting transition temperature T-cmax on the on-site effective Coulomb interaction U using the constrained random-phase approximation. We focus on the commonly used one-band model of the cuprates, including only the antibonding combination of the Cu d(x2-y2) and O p(x) and p(y) orbitals and find a screening-dependent trend between the static value of U and T-cmax for the parent compounds of a large number of hole-doped cuprates. Our results suggest that superconductivity may be favored by a large on-site Coulomb repulsion. We analyze both the trend in the static value of U and its frequency dependence in detail and, by comparing our results to other works, speculate on the mechanisms behind the trend.
Excitation of image states is predicted to occur frequently in inverse-photoemission spectroscopy due to diffracted electron beams traveling almost parallel to the surface. A simple kinematic model, as well as full calculations using inverse low-energy electron-diffraction theory, show the effect in the case of Cd(0001).
The electronic structure for a metallic superlattice system, Mo/V, has been calculated using the linear muffin-tin orbital method in the atomic-sphere approximation (ASA). Total energies have been calculated in the local-density approximation (LDA). Emphasis has been given to the electronic-density variation in these materials in order to understand hydrogen storage in metallic superlattices. It is found, within effective-medium theory, that the balance between charge transfer at the interfaces and volume changes of the constituents determines preferred hydrogen sites in Mo/V. Total-energy considerations favor a tetragonal distortion of the Mo/V superlattice over a cubic structure. Results for the LDA bands, charge transfer, and density of states are compared to other electronic-structure calculations of metallic superlattices, which gives rise to a consistent picture of electronic effects in these systems. The dependence of the results on the ASA is discussed. © 1992 The American Physical Society.
Orbital magnetization is known empirically to play an important role in several magnetic phenomena, suchas permanent magnetism and ferromagnetic superconductivity. Within the recently developed “modern theoryof orbital magnetization,” theoretical insight has been gained into the nature of this often neglected contributionto magnetism but is based on an underlying mean-field approximation. From this theory, a few treatments haveemerged which also take into account correlations beyond the mean-field approximation. Here, we apply thes cheme developed in a previous work [F. Aryasetiawan et al., Phys. Rev. B 93, 161104(R) (2016)] to thespin- 1/2 Haldane-Hubbard model to investigate the effect of charge fluctuations on the orbital magnetizationwithin the GW approximation. Qualitatively, we are led to distinguish between two quite different situations:(i) When the lattice potential is larger than the nearest-neighbor hopping, the correlations are found to boostthe orbital magnetization. (ii) If the nearest-neighbor hopping is instead larger than the lattice potential, thecorrelations reduce the magnetization. The boost and reduction are identified to stem from interband andintraband correlations, respectively, and the relative importance of the two varies with the strength of the latticepotential. We finally study graphene with parameters obtained from first principles.
We consider two commonly used approaches for the ab initio calculation of optical-absorption spectra, namely, many-body perturbation theory based on Green’s functions and time-dependent density-functional theory (TDDFT). The former leads to the two-particle Bethe-Salpeter equation that contains a screened electron-hole interaction. We approximate this interaction in various ways, and discuss in particular the results obtained for a local contact potential. This, in fact, allows us to straightforwardly make the link to the TDDFT approach, and to discuss the exchange-correlation kernel fxc that corresponds to the contact exciton. Our main results, illustrated in the examples of bulk silicon, GaAs, argon, and LiF, are the following. (i) The simple contact exciton model, used on top of an ab initio calculated band structure, yields reasonable absorption spectra. (ii) Qualitatively extremely different fxc can be derived approximatively from the same Bethe-Salpeter equation. These kernels can however yield very similar spectra. (iii) A static fxc, both with or without a long-range component, can create transitions in the quasiparticle gap. To the best of our knowledge, this is the first time that TDDFT has been shown to be able to reproduce bound excitons.
In photoemission spectra of strongly correlated systems one usually observes a satellite structure below the main peak. Description of such satellite structures in the commonly used GW approximation has been found to be insufficient. To account for these satellite structures that originate from short-range correlations, we have developed a T-matrix formalism for performing ab initio calculations on real systems. The method is applied to Ni and we obtain a satellite structure below the Fermi level as well as a reduced exchange splitting. We also found a new interesting satellite structure above the Fermi level, which can be ascribed to particle-particle scattering.
Resonant in situ photoemission from Mn 3d states in Ga(1−x)MnxAs is reported for Mn concentrations down to the very dilute level of 0.1%. Concentration-dependent spectral features are analyzed on the basis of first-principles calculations for systems with selected impurity positions as well as for random alloys. Effects of direct Mn-Mn interaction are found for concentrations as low as 2.5%, and are ascribed to statistical (nonuniform) distribution of Mn atoms.
The energy-resolved momentum densities of thin polycrystalline aluminium films have been measured using electron momentum spectroscopy (EMS), for both the valence band and the outer core levels. The spectrometer used for these measurements has energy and momentum resolutions of around 1.0 eV and 0.15 atomic units, respectively. These measurements should, in principle, describe the electronic structure of the film very quantitatively, i.e. the dispersion and the intensity can be compared directly with theoretical spectral momentum densities for both the valence band and the outer core levels. Multiple scattering is found to hamper the interpretation somewhat. The core-level intensity distribution was studied with the main purpose of setting upper bounds on these multiple-scattering effects. Using this information we wish to obtain a full understanding of the valence band spectra using different theoretical models of the spectral function. These theoretical models differ significantly and only the cumulant expansion calculation that takes the crystal lattice into account seems to describe the data reasonably well.